CQSRG Online Educational Material
How to locate an earthquake
CQSRG (pronounced CQ Surge) has been researching the earthquake seismicity of Eastern Central Queensland since it began operation in 2002.
There are several methods used to work out where an earthquake occurred. In all cases the starting point is to obtain reliable recordings of the energy waves, created by the earthquake, that travel through the ground. These seismic waves may be recorded on a number of stations at a distance from the point on the Earth's surface under which the earthquake occurred. This point on the Earth's surface is called the earthquake's epicentre.
Of course, the place where all earthquakes happen is not actually on the Earth's surface; it is invariably at some distance under the surface - usually several kilometres under the surface. The place down under the epicentre where the earthquake erruppted is called the hypercentre or focus of the earthquake.
Whilst it is relatively easy to determine an earthquake's epicentre, it can be quite difficult to determine the hypercentre.
A seismogram is a graphical display of the very small ground movement caused by the radiation of seismic energy outwards from an earthquake's hypercentre in waves.
The vertical axis represents the ground movement; which may be measured as displacement, velocity, or acceleration. In the graph on the right the ground movement is measures in digital count units that represent the velocity of the ground movement in metres per second.
The horizontal axis represents the time that has elapsed since the start of the recording, in seconds.
It is clear from examining the graph that at different times the ground velocity has different values. We see, for instance, that prior to the arrival of the first seismic wave the graph is rather flat. This is the normal background ground movement. Then, quite suddenly, the ground velocity increases. It peaks somewhat, and then continues for several seconds at an elevated but moderate level, before again increasing in level rather suddenly and gradually increasing to a maximum level, and then gradually decreasing back towards the background level.
When an earthquake suddenly ruptures the Earth's crust at the focus it is like a large explosion. This explosion generates a combination of pressure and shear waves in the crust that radiate outwards from the focus.
The pressure waves are longitudinal waves in many ways are similar to sound waves - except that sound waves travel through the air, whereas the longitudinal seismic waves travel through the Earth's crust.
The shear waves are traverse waves. They are somewhat similar to the waves you see when you throw a stone into a pond - except that they are not travelling over the surface. They are travelling through the body of the Earth's crust.
The compressional waves travel faster than the shear waves, and will arrive at a distant measuring station first. They are therefore called the Primary or P waves.
The shear waves arrive at a distant measuring station after the compressional waves. They are therefore called the Secondary or S waves.
The graph on the right is the same seismogram as we saw previously; but, in this image, we see that the arrival times of the P and S waves are highlighted. It also shows that the time difference between the P and S arrivals is (in this case) 14.12 seconds.
As well as the Primary and Secondary seismic waves an earthquake generates a number of other types of seismic waves that we will not be discussing here (because thay are not used to locate earthquakes). These other seismic waves travel slower than the P and S waves; and, in fact, if you look closely at the graph after the S arrival, you will see a sequence of waves that are more widely spaced than the waves between the S and P arrivals. Some of these waves are what are called surface waves and an indication that this particular earthquake was particularly shallow - probably within a couple of kilometres of the Earth's surface.
The further the waves travel away from the earthquake's focus the longer is the time difference between the arrival of the P and S wave fronts. So, for measurement stations placed at different distances from the earthquake epicentre, we will get seismogram records from which different S-P times can be measured.
It turns out that the relationship between the measured S-P time and the distance of the measurement station from the earthquake epicentre is stable enough for it to be used to calculate the approximate distance from the earthquake epicentre to the measurement station.
In Continental Crust the distance from an earthquake epicentre to a station that measures the arrival times of the P and S wave fronts can be calculated using the following approximation:
Distance(km) ~ (S-P) * 7.9
For instance, in the seismograph above, the S-P time was measured as being 14.12 seconds. Therefore the station that recorded the arrival of the waves was situated about 112 km from the earthquake's epicentre.
In order to map the distance from the measurement station to the earthquake epicentre we need to know where exactly the measurement station is located.
We see from the graphs above that the name used to designate the measurement station is YNG.
The International Seismological Centre (ISC) in the England maintains an online register of all official seismic monitoring stations in the World.
By searching the ISC Station Register I get the following information about the YNG station.
Code Station name/Region Latitude Longitude Elevation Depth Prime Status ============================================================================================================== YNG Young -34.29800 148.39631 460.0 Open New South Wales,Australia
This gives us the Latitude and Longitude of the station - which can be used to accurately locate the station on Google Earth Pro. Also using the features of Google earth Pro we can draw a circle of radius 112 km centred on the YNG seismic station.
This indicates that the earthquake epicentre is somewhere on the circumference of that circle.
Ok... now ... what if we had S-P times for another, different, monitoring station? That would allow us to draw another circle around that station - and it will intersect the circle drawn about the YNG station in (at most) two places! That would restrict the location of the epicentre to one of two possible places.
If we get data from a third monitoring station we can draw a third circle, that will restrict the location of the epicentre even further to a very small area.
The more stations we can get data from the more circles we can draw to confine the epicentre location more and more.
We don't actually need the seismograms themselves from which to measure the P and S arrival times. Often various Seismology Agencies publish tables of station arrival times. The data in those tables can be used to calculate the S-P times for the various stations. We then use the ISC online register to locate the stations. Then we can use Google earth Pro to map the station to epicentre distances, and hence to locate the earthquake's epicentre.
The following table provides data from three stations that recorded the earthquake shown in the above seismograms.
The S-P times can be calculated by subtracting the P arrival time from the S arrival time.
The latitudes and longitudes for each of the stations can be found by searching the online registerISC station register.
We can see that the distance circles for the three stations all intersect at one place (approximately).
By focusing in on the intersection location we see that, in fact, the three circles do not intersect in exactly the same place.
At the intersection location the three circles form a triangle. This is called the error triangle, and the most likely location of the earthquake epicentre is somewhere inside that triangle.
The epicentre is located about 30 km south of the Orange township in New South Wales; and about 20 km south east of the Cadia Valley Gold Mine (the largest underground gold minning operations in Australia).
It was noted earlier that the seismograms for this earthquake exhibited very prominent surface waves; indicating that the event was very shallow. Given that the vast majority of natural earthquakes in Australia usually occur about 10 km beneath the surface, and do not produce any significant surface waves, the close presence of a large mining operation to the located epicentre must cause speculation that this was not, indeed, a natural earthquake. Further investigation would be warranted.
The table below is a typical example of such data (note that this is data from a different earthquake to the one above).
The GA data table actually makes it very easy to map the distances - because they include their distance calculations in the table. These distances will be more accurate than those calculated using the formula presented above. Perhaps you can use the GA data to determine a more accurate factor than 7.9.
The Geoscience Australia (GA) also allows you to get an RSS feed that sends out email notifications whenever it locates an earthquake. This means that you can practice locating earthquakes using the GA published data, and also check whether your location is correct by comparing it with their map.